STATE OF

GLOBAL AIR/2018
A SPECIAL REPORT ON GLOBAL EXPOSURE TO AIR POLLUTION
AND ITS DISEASE BURDEN

The State ofTheGlobal

StateAir
of Global
is a collaboration
Air is a collaboration
between the
between
Healththe
Effects Institute
Institute
and the for Health
Institute Metrics
for Health and and
Metrics Evaluation’s Global
Evaluation’s Burden
Global of Disease
Burden Project
of Disease Project.
i STATE OF GLOBAL AI R / 2 0 1 8 andofthe
Citation: Health Effects Institute. 2018. State Health
Global AirEffects
2018. Institute.
Special Report. Boston, MA:Health Effects Institute.
What is the State of Global Air?
The State of Global Air report brings into one place the most
recent information available on levels and trends in air quality
and health for countries around the globe. This year we focus
not only on ambient (outdoor) air pollution but also, for the
first time, on household air pollution from the burning of solid
fuels for cooking and heating, a major contributor to pollution
both inside and outside the home.

Who is it for?
The report is designed to introduce citizens, journalists, policy
makers, and scientists to efforts to estimate and track human
exposure to outdoor and household air pollution and their
impacts on health as part of the comprehensive Global Burden
of Disease project.

How can I explore the data?

This report has a companion interactive website, which
provides the tools to explore, compare, and download data
tables and graphics with the latest outdoor and household
air pollution levels and associated burden of disease. These
data are available for individual countries and geographic and
economic regions, as well as for highlighting trends from 1990
to 2016.

Where will I find information on:

The Global Burden of Disease and
State of Global Air projects ...........................................page 1
Exposure to Ambient Air Pollution:
Levels and Trends...........................................................page 3
Exposure to Household Air Pollution from Burning
of Solid Fuels: Levels and Trends...................................page 6
Burden of Disease Attributable to Air Pollution.............page 9
The Key to Air Quality Progress: Understanding
the Major Sources of Air Pollution...............................page 16
Numbers, Numbers Everywhere...................................page 19
INTRODUCTION

S
tudies from across the world have documented the many
Fine particle air pollution is the largest environmental
ways in which air pollution can affect people’s health, in-
risk factor worldwide, responsible for a substantially
cluding making it difficult to breathe for those with asth-
larger number of attributable deaths than other more
ma or other respiratory diseases, sending the young and
well-known behavioral risk factors such as alcohol
old to hospital or causing them to miss school or work,
use, physical inactivity, or high sodium intake.
and contributing to early death from heart and lung dis-
ease. New studies continue to broaden our understanding of the wide
equal to 2.5 micrometers in aerodynamic diameter, or PM2.5), one compo-
range of effects that air pollution can have on human health.
nent of air pollution, ranked as the 6th-highest risk factor for early death
This State of Global Air 2018 report presents the latest analysis of
(see Figure 1). Worldwide exposure to PM2.5 contributed to 4.1 million
worldwide air pollution exposures and health impacts. It draws from
deaths from heart disease and stroke, lung cancer, chronic lung disease,
the most recent evidence (produced in 2016) as part of the Global
and respiratory infections in 2016. PM2.5 was responsible for a substan-
Burden of Disease (GBD) project of the Institute for Health Metrics
tially larger number of attributable deaths than other more well-known
and Evaluation (IHME) (published in 2017; see Additional Resources
at the end of this report). risk factors (such as alcohol use, physical inactivity, or high sodium in-
take) and for an equivalent number of attributable deaths as high cho-
As it did last year, the report offers a global update on outdoor, or
lesterol and high body mass index. Ozone, another important component
ambient, air pollution. The most recent GBD analysis has continued to
of outdoor air pollution, whose levels are on the rise around the world,
identify ambient air pollution as one of the most important risk factors
contributed to 234,000 deaths from chronic lung disease.
contributing to death and disability (see “Defining Ambient Air Pollution”
However, the GBD initiative has also documented that millions of
textbox). Ambient particulate matter (particulate matter less than or
people around the world are exposed to
Figure 1. Global ranking of risk factors by total number of deaths from all air pollution in their homes arising from
causes for all ages and both sexes in 2016. the use of solid fuels (e.g., coal, wood,
and dung) for cooking and heating. The
GBD 2016 analysis estimates that ex-
posure to “household air pollution” also
has a substantial impact on health and is
ranked 8th in risk factors for early death,
with 2.6 million attributable deaths in
2016. Both individually and collectively,
ambient air pollution and household air
pollution impose a substantial burden on
public health.
To offer a more comprehensive per-
spective on air pollution, this year’s State
of Global Air 2018 provides additional
data on:

• Household air pollution — global,

regional, and country-level estimates
of the proportion of populations
exposed to household air pollution
and of the public health burden
Explore the rankings further at the IHME/GBD Compare site. attributable to these exposures, and

1 STATE OF GLOBAL AI R / 2 0 1 8
 • T otal air pollution — global, regional, and country-level
estimates of the total public health burden attributable to
ambient air pollution (PM2.5 and ozone) and to household air
pollution.
This year’s report also adds a discussion on the identification of
the major sources of air pollution, focusing on recent work in both
India and China, and the implications these sources have for the
actions necessary to reduce emissions and ultimately public health
burden from air pollution.
The State of Global Air website offers interactive features to ex-
plore and compare visually ambient, household, and total air pollu-
tion levels and their health impacts in individual countries and at the
regional level around the world.

Defining Ambient Air Pollution increases in mortality from respiratory dis-

Air pollution is a complex mixture of gases
and particles whose sources and composition
vary spatially and temporally. While hundreds
of different chemical compounds can be mea-
sured in air, governments typically measure
only a small subset of gases and particles
as indicators of the different types of air
pollution and the different types of major
sources contributing to the pollution. PM2.5
and tropospheric ozone (i.e., ozone found in
the atmosphere nearest the earth, where we
live and breathe) are the two indicators used
to quantify exposure to outdoor, or ambient,
air pollution in the GBD project.
PM2.5, defined as fine particles with
aerodynamic diameters less than or equal to
2.5 micrometers, is the most consistent and
robust predictor of mortality from cardio-
vascular, respiratory, and other diseases in
studies of long-term exposure to air pollu-
tion. Long term is defined by annual average
exposures over several years. PM exposure
is measured in micrograms per cubic meter
(µg/m3). Ozone, a harmful gas produced
via the atmospheric reactions of precursor
emissions, has itself been implicated in
ease. When describing exposures to ozone,
scientists focus on seasonal, rather than
annual, average concentrations because
ozone levels are higher in the warm season
in the mid-latitudes where most epidemio-
logical studies have been conducted. Ozone
exposure is measured in parts per billion
(ppb). Exposure to each pollutant is repre-
sented by population-weighted averages,
which take into account the proportions of
the population living in areas with different
levels of pollution.
How Are Ambient Air Pollution
Levels Estimated Around the World?
Although many high-income countries around
the world operate extensive networks of air
quality monitoring stations in urban areas,
providing continuous hourly measurements
of pollution levels each day, this is not the
case for most countries. These ground-level
measurements of air quality have been
the basis for most studies of the potential
health effects of air pollution and air quality
management. However, other approaches are
needed to provide a consistent view of air
pollution levels throughout the world,
including in the many rapidly developing
urban areas of low- and middle-income
countries and in the large rural and suburban
areas that lack any air quality monitoring
stations. For these areas, scientists rely
on air quality observations from satellites,
combined with information from global
chemical transport models and available
ground measurements, to estimate global
annual average exposure to PM2.5 system-
atically, beginning in blocks, or grid cells,
covering 0.1° × 0.1° of longitude and latitude
(approximately 11 km × 11 km at the equator).
Taking into account the population in each
block within a country, scientists then
aggregate the estimated exposure concentra-
tions to national-level population-weighted
averages for a given year. The GBD analysis
was conducted in 2016 using data from 1990
to 2016, the most recent year for which the
necessary data were available. For ozone, a
global chemical transport model was used to
calculate a seasonal (summer) average for
each grid cell, while accounting for variation
in the timing of the ozone season in different
parts of the world. The process for estimating
national-level population-weighted average
ozone exposures was the same.

2 STATE OF GLOBAL AI R / 2 0 1 8
EXPOSURE TO AMBIENT AIR POLLUTION: LEVELS AND TRENDS
AMBIENT PM 2.5 CONCENTRATIONS AROUND THE WORLD
A
round the world, ambient levels of PM2.5 continue to ex-
ceed the Air Quality Guideline established by the World
Health Organization (WHO). WHO set the Air Quality
Guideline for annual average PM2.5 concentration at
10 µg/m3 based on evidence of health effects of long-
term exposure to PM2.5 but acknowledged that it could
not rule out health effects below that level. For regions of the world
where air pollution is highest, WHO suggested three interim targets
set at progressively lower concentrations: 35 µg/m3, 25 µg/m3, and
15 µg/m3. Figure 2 shows where these guidelines were still exceeded
in 2016.
Based on these data and knowledge of the populations in each
country for 2016, 95% of the world’s population lived in areas that
exceeded the WHO Guideline for PM2.5. Fifty-eight percent of the
global population resided in areas with PM2.5 concentrations above
the WHO Interim Target 1 (IT-1, 35 µg/m3); 69% lived in areas exceed-
ing IT-2 (25 µg/m3), and 85% lived in areas exceeding IT-3 (15 µg/m3).
The highest concentrations of population-weighted annual av-
erage PM2.5 (see “Defining Ambient Air Pollution” textbox) in 2016
were in countries in North Africa (e.g., Niger at 204 µg/m3 and Egypt
at 126 µg/m3), West Africa (e.g., Cameroon at 140 µg/m3 and Nigeria
Figure 2. Comparison of 2016 annual average PM2.5 concentrations to the WHO Air Quality Guideline.
3 STATE OF GLOBAL AI R / 2 0 1 8
Ninety-five percent of the world’s population lives
in areas exceeding WHO Guideline for healthy air.
Nearly 60% lives in areas that do not meet even the
least-stringent air quality target from WHO.
at 122 µg/m3), and in the Middle East (e.g., Saudi Arabia at 188 µg/
m3 and Qatar at 148 µg/m3). The high outdoor concentrations in these
regions were due mainly to windblown mineral dust. However, in
some of these countries (Niger, Nigeria, and Cameroon), high pro-
portions of the population burn solid fuels in the home and may also
engage in open burning of agricultural lands or forests, both of which
can also contribute substantially to outdoor air pollution.
The next-highest concentrations appeared in South Asia where
combustion emissions from multiple sources, including household solid
fuel use, coal-fired power plants, agricultural and other open burning,
and industrial and transportation-related sources, are the main con-
tributors. The population-weighted annual average PM2.5 concentra-
tions were 101 µg/m3 in Bangladesh, 78 µg/m3 in Nepal, and 76 µg/m3
in both India and Pakistan. The population-weighted annual average
concentration in China was 56 µg/m3. Estimates for population-weight-
ed annual average PM2.5 concentrations were lowest (≤ 8 µg/m3) in
Australia, Brunei, Canada, Estonia, Finland, Greenland, Iceland, New
Zealand, Sweden, and several Pacific island nations.
Figure 3. Trends in population-weighted annual average PM2.5
concentrations in the 10 most populous countries plus the European
Union, 2010–2016.
Explore the trends in other countries and regions at the State of Global Air interactive site.
MAIN TRENDS IN PM CONCENTRATIONS
2.5
Global population-weighted PM2.5 concentrations increased by 18%
from 2010 (43.2 µg/m3) to 2016 (51.1 µg/m3). As the previous discus-
sion on the most recent country data suggests, the global level of
population-weighted PM2.5 is influenced strongly by the levels of air
pollution in populous regions and countries.
China’s air pollution exposures have stabilized
and even begun to decline slightly; Pakistan,
Bangladesh, and India, in contrast, have experienced
the steepest increases in air pollution levels since
2010.
Figure 3 illustrates the trends in population-weighted PM2.5
concentrations for the 10 most highly populated countries in the
world along with the European Union from 2010 to 2016 (see the
interactive site for all years 1990 to 2016). India, Bangladesh, Pa-
kistan, and China have all experienced both high concentrations
and increasing trends in PM2.5 exposure, but there are noteworthy
distinctions. Although China experienced substantial increases in
population-weighted exposures before 2010 — reflecting in part
the dramatic scale of economic development in recent decades —
since then the exposures have stabilized and even begun to decline
4 STATE OF GLOBAL AI R / 2 0 1 8
slightly. Pakistan, Bangladesh, and India,
on the other hand, have experienced the
steepest increases in air pollution levels
since 2010 and now present the highest
sustained PM2.5 concentrations among
the countries shown here.
While Saharan Desert dust events
are common and annually affect North
Africa, between 2015 and 2016 anom-
alous wind patterns led to major dust
events that also affected highly popu-
lated regions of West Africa. Nigeria,
in particular, saw dramatic increases in
PM2.5 concentrations resulting from an
extensive dust storm in late 2015 and
early 2016. However, longer-term trends
suggest declines in PM2.5 exposures in
Nigeria over the last 26 years, with limit-
ed evidence pointing to general declines
in mineral dust emissions and open burn-
ing. Concentrations in the other highly
populated countries (Russia, Indonesia,
Japan, Brazil, and the United States, as
well as the European Union) declined
since 1990, yet, with the exception of
the United States, remain above the WHO Guideline value.
Excluding the recent evidence of stabilization and slight declines
in population-weighted concentrations in China, the disparities
among these large countries have grown substantially over time.
Less-polluted locations have become cleaner, while PM2.5 concen-
trations have increased in the more polluted locations, especially in
South Asia. As a result, what was a 6-fold range in 1990 in pop-
ulation-weighted average concentrations among these countries
(excluding Nigeria, given the likelihood of the 2015–2016 increase
being transient) increased to an 11-fold range in 2016.
HOW AMBIENT OZONE CONCENTRATIONS VARY
AROUND THE WORLD
The global ozone map (Figure 4) indicates that seasonal popula-
tion-weighted average ambient ozone concentrations generally vary
less around the world compared with PM2.5. Ozone concentrations
were relatively higher in the United States, West and Central sub-Sa-
haran Africa, and throughout the Mediterranean, the Middle East,
South Asia, and China. Globally, population-weighted ozone con-
centrations have increased, which reflects a combination of factors,
including increased emissions of ozone precursors (such as nitrogen
oxides) coupled with warmer temperatures, especially at mid-lati-
tudes in rapidly developing economies. Ozone concentrations are led
by ongoing increases experienced in China, India, Pakistan, Bangla-
desh, and Brazil.
Ozone levels remain high across both higher-income
and lower/middle income regions of the world.

Figure 4. Population-weighted seasonal average ozone concentrations in 2016.

Explore the data on the State of Global Air interactive site. For country abbreviations, see ISO3 website.

5 STATE OF GLOBAL AI R / 2 0 1 8
EXPOSURE TO HOUSEHOLD AIR POLLUTION FROM BURNING OF SOLID
FUELS: LEVELS AND TRENDS
H
ousehold air pollution from the use of solid fuels for
cooking and heating is the 8th leading global risk factor
contributing to disease burden (see Figure 1). System-
atic measurements of PM2.5 in households using solid
fuels around the world are not available, but numerous
individual studies indicate that household air pollution
concentrations often far exceed those observed in ambient air. For
example, a recent review of such studies reported that, in homes
where no interventions had been undertaken to reduce exposures
from the burning of solid fuels (e.g., by providing improved biomass
stoves with and without chimneys or cleaner fuels such as LPG or
ethanol), average exposures were as high as 220 µg/m3 — approxi-
mately 6 times higher than the WHO Level 1 Interim Target of 35 µg/
m3. Even after efforts to reduce exposures using a range of interven-
tions, this review reported that average in-home exposures were still
approximately 100 µg/m3, or 2.8 times higher than the same WHO
interim target (see Additional Resources).
The GBD project estimates exposure to household air pollution
by combining information from various surveys (including approxi-
mately 680 studies from 150 countries for the period 1980 to 2016),
Estimating Global Exposures to Household Air
Pollution
Estimation of exposure to household air pollution for the IHME
GBD initiative begins by determining the proportion of households
using solid fuels of any kind rather than gas or electric sources
for cooking. Data on fuel use were extracted from numerous
surveys (demographic and health surveys, living standards mea-
surement surveys, multiple indicator cluster surveys, and world
health surveys, as well as country-specific sources and the WHO
household energy database). In 2016, IHME extracted 680 data
points from 150 countries from these resources. These data are
used together with demographic data to estimate the proportions
of the population (stratified by age and sex) exposed to emissions
from household use of solid fuel for each national and, where
available, subnational geographic area in the GBD. The proportion
of the total population exposed to household solid fuels is used in
this report as an estimate of household air pollution exposure.
Subsequently, in order to apply the integrated exposure–
response functions to estimate disease burden, the use of solid
fuels is translated into indoor PM2.5 concentrations and then into
6 STATE OF GLOBAL AI R / 2 0 1 8
People living in homes using solid fuels can face
PM2.5 levels 6 times higher than even the least-
stringent WHO Interim Air Quality Target for PM2.5
and as much as 20 times higher than the full WHO Air
Quality Guideline of 10 µg/m3.
primarily at the national or regional level, to calculate the proportion
of households using solid fuels. These data are then modeled to pro-
vide complete trends and estimates for all countries. Combining this
information with population data for each location allows for estima-
tion of the proportion of the population (grouped by age and sex) that
is using solid fuels and that is assumed to be exposed to household
air pollution (see textbox “Estimating Global Exposures to Household
Air Pollution”). The State of Global Air presents these proportions
of populations using solid fuels (all ages and both sexes combined)
for the world as a whole, for different geographic regions, and for
individual countries. The map in Figure 5 displays the proportion of
population exposed to household air pollution from burning solid
fuels for each country across the globe. Details for individual coun-
tries can be explored on the interactive website.
exposures for men, women, and children. This mapping relies on
indoor measurements of household-air-pollution–related PM2.5
from 90 studies in 16 countries. For the GBD 2016 analysis, a
linear model was used to estimate the indoor PM2.5 concentra-
tions based on IHME’s sociodemographic index (a metric assigned
to every country in the world based on average income per
person, educational attainment, and total fertility rate). Covari-
ates were included in the model to indicate whether measured
concentrations were made in the kitchen (or elsewhere in the
home) or were measurements of personal exposure, as well as
the measurement duration. The relationship described by the
linear model was then used to predict indoor concentrations for
all geographic locations and years. Finally, these indoor concen-
trations were translated to exposures by applying the ratio of
personal exposures to indoor concentrations based on a subset of
seven studies from six countries that included paired personal and
indoor measurements. These ratios were modeled separately for
men, women, and children based on available differences in the
time spent in household activities that would involve exposure to
household air pollution.
For more details on these methods, see references in the
Additional Resources section.
Figure 5. Proportion of population exposed to household air pollution from burning of solid fuels in 2016.

Explore the data on the State of Global Air interactive site. For country abbreviations, see ISO3 website.

den of Disease from Major Air Pollution Sources (GBD MAPS) project
In 2016, a total of 2.45 billion people — one in three
has quantified the contribution of household burning of solid fuels to
global citizens — were exposed to household air
ambient PM2.5 levels and the associated disease burden in China and
pollution from the use of solid fuels.
India. In China, household burning of biomass and coal contributed
In 2016, a total of 2.45 billion people (33.7% of the global popu-
Figure 6. Number of people and percentage of population
lation) were exposed to household air pollution. Figure 5 illustrates exposed to household air pollution from solid fuel burning in
wide disparities in the exposures to household air pollution around countries with populations over 50 million and at least 10%
the world, with a clustering of two dozen countries in Africa that solid fuel use in 2016. Countries listed in order of absolute
had over 90% of their populations exposed to household air pollu- number of people exposed.
tion from the use of solid fuels. Ethiopia, Democratic Republic of the
Congo, and Tanzania each had 96% or more of their populations ex-
posed to household air pollution. Figure 6 ranks the 13 countries with
populations of over 50 million that also have more than 10% of their
populations exposed to solid fuels. Even where the percentage of
population exposed to household solid fuel use is lower, the numbers
of people potentially exposed can be substantial. India and China
— with 43% and 30%, respectively, of their populations using solid
fuels — had the largest numbers of people exposed to household air
pollution in 2016: 560 million in India and 416 million in China.
The GBD project treats household use of solid fuels as a sepa-
rate risk factor from outdoor air pollution. However, household air
pollution can be an important contributor to outdoor air pollution,
although the extent of this contribution varies by location and time
and has not been estimated for most countries. The HEI Global Bur- Explore the data on the State of Global Air interactive site.

7 STATE OF GLOBAL AI R / 2 0 1 8
about 19% of total population-weighted PM2.5 con-
centrations; in India, household burning of biomass
was responsible for about 24% of PM2.5 levels (see
“The Key to Air Quality Progress: Understanding the
Major Sources of Air Pollution” later in this report).
Figure 7. Trends in proportion of population exposed to household
air pollution from burning of solid fuels for selected regions of the
world.

MAIN TRENDS IN HOUSEHOLD AIR

POLLUTION EXPOSURE
Trends in population exposure to household air pol-
lution are estimated in the GBD project by evaluat-
ing trends in the proportion of the households using
solid fuels. Currently there are insufficient data to
estimate accurately changes in exposure directly,
given differences in types of fuels and appliances,
house design, and cooking practices between and
within countries. Trends show that the proportion of
the population using solid fuels in major world re-
Explore the data on the State of Global Air interactive site.
gions (Figure 7) has been declining in many regions
of the world over the last 26 years (1990–2016).
Some of the most dramatic declines have occurred Figure 8. Trends in proportion of population exposed to household
in Asia and southern Africa and throughout most air pollution from burning of solid fuels by selected GBD
of South America. However, in East, Central, and sociodemographic index (SDI) regions.
West sub-Saharan Africa, where the percentages
of population exposed to household air pollution
have been the highest, there has been little change
in exposure.
Progress toward reductions in reliance on
burning solid fuels in the home has been fast-
est in middle- and high-income countries, and
slowest in the rural areas of low- and middle-in-
come regions of the world. These trends are il-
lustrated in Figure 8, which groups populations
in the geographic regions in the previous figure
according to where they fall on the development
spectrum, using the GBD’s sociodemographic in-
dex (SDI)*, with the low SDI group comprised
primarily of African countries. (See GBD Socio-
Demographic Income Regions.) Explore the data on the State of Global Air interactive site.

Despite global population growth over the

past 26 years, there has been a net decline in the total numbers
of people relying on solid fuels from just over 3 billion in 1990 to
about 2.4 billion in 2016. Decline or stabilization of these num-
bers is evident even in many of the largest countries (>50 million
population) with substantial proportions of households relying on
solid fuels. China, for example, has reduced the estimated popula-
tion relying on solid fuels to about 416 million in 2016, down from
996 million in 1990. Given the combination of population growth
and limited changes in the percentages of their populations us-
ing solid fuels, Nigeria, Ethiopia, the Democratic Republic of
Congo, and Tanzania have all seen net increases in populations ex-
posed to household air pollution.
* The sociodemographic index (SDI) is the GBD’s numerical index of develop-
ment that reflects underlying social, demographic, and economic differences
among populations in countries or in other geographic areas.

8 STATE OF GLOBAL AI R / 2 0 1 8
BURDEN OF DISEASE ATTRIBUTABLE TO AIR POLLUTION
W
hat do we mean by “burden of disease” and how is
it measured in the GBD project? The GBD initiative
measures the burden of disease in terms of (1) the
numbers and age-standardized rates of deaths in
a given year and (2) the numbers of healthy years
of life lost from death or disability, represented by
“disability-adjusted life-years” (DALYs) (see textbox “Understanding
the Burden of Disease: Deaths and DALYs”). For each risk factor con-
sidered by GBD, such as air pollution, these measures of disease
burden are estimated for each country or subnational area from 1990
to 2016.
Understanding the Burden of Disease: Deaths and
DALYs
The burden of disease due to exposure to a pollutant is calcu-
lated using an estimation of the number of deaths and disabili-
ty-adjusted life-years (DALYs). The number of deaths attributable
to air pollution in a given year includes deaths that have likely
occurred months or even years earlier than might be expected in
the absence of air pollution (as in the case of a child dying from a
lower-respiratory infection [LRI] ). DALYs provide an overall mea-
sure of the loss of healthy life expectancy and are calculated as
the sum of the years of life lost from a premature death and the
years lived with disability (e.g., blindness caused by the disease
diabetes). An important insight gained by using DALYs rather
than just the numbers of deaths is that DALYs account for the
age at which disease or death occurs. For example, air pollution
9 STATE OF GLOBAL AI R / 2 0 1 8
Estimating the burden of disease attributable to air pollution in
geographic regions over time requires three major components: (1)
estimates of population-weighted average annual exposure to PM2.5,
ozone, and household air pollution, as described earlier in this report
(which are compared against a “minimum risk exposure level,” an
exposure level below which there is no evidence of additional risk of
mortality or disease); (2) mathematical functions, derived from epi-
demiological studies, that relate the different levels of exposure to
the age, sex, and cause-specific health impacts (e.g., mortality from
heart disease and stroke, and incidence of lung cancer); and (3) spa-
tially and temporally resolved estimates of the underlying number of
deaths and DALYs for each of the causes of death or disability linked
to air pollution. Sources for more information on these methods are
listed in “Additional Resources” at the end of this report.
An extensive scientific literature has documented a
wide range of air pollution health effects, including
asthma, increased hospitalizations, illness, and
reduced life expectancy from heart and lung disease.
What are air pollution’s effects on health? An extensive scientific
literature has documented a wide range of health effects: short-term
effects such as worsening of asthma and increased hospitalizations
for various cardiovascular and respiratory diseases on high-pollution
days, as well as increased mortality and illness from cardiovascular
contributes to LRIs in children, but the number of deaths is small
relative to the numbers of air pollution–related deaths from heart
disease, which tend to occur in older adults. However, because
children who die from LRIs have lost many more years of healthy
life, this burden is appropriately reflected in a larger number of
DALYs.
Burden is also measured in terms of age-standardized death
rates and DALY rates (i.e., the number of deaths or DALYs per
100,000 people). Age-standardized rates are important because
they adjust for population size and the age structure of each
country’s population. This means that the rates in two countries
can be compared as if the countries had the same population
characteristics. Otherwise, in a country with a large and older
population, the total number of deaths attributable to air pol-
lution would be larger than that in a country with a smaller or
younger population, even if exposure levels were the same.
 The Growing Burden of Disease Due to Ambient Air Pollution Exposure in an Aging World
The latest estimates for the global burden
of disease due to ambient PM2.5 air pollu-
tion make clear that it takes its greatest
toll in the elderly and middle-aged, in par-
ticular from non-communicable diseases
(cardiovascular disease, stroke, lung can-
cer, and COPD). All told, in 2016, ambient
PM2.5 air pollution contributed to 1.8 million
deaths and 22 million DALYs in those older
than 70 years and 1.3 million deaths and
37 million DALYs in those between 50 and
69 years old from these diseases.
The growing importance of these
non-communicable diseases in the burden
of disease for air pollution exposure reflects
worldwide increases in life expectancy at
birth over the last 50 years and improve-
ments in prevention and treatment for
communicable diseases (malaria, pneumo-
nia, etc.). The result has been an epidemio-
logical transition from historical patterns in
which the burden of disease was dominated
by communicable diseases that resulted in
deaths and disability in the very young to
patterns in which the burden is dominated
by non-communicable diseases in those
Percentage of total DALYs resulting from non-communicable diseases by country for both sexes
and all ages in 2016.
10 STATE OF GLOBAL AI R / 2 0 1 8
over age 50. The textbox figure shows that
in much of the world non-communicable
diseases dominate the burden of disease: of
some 1.5 billion DALYs in 2016, comprising
39 million deaths and 648 million years
lived with disability, more than 80% were in
those 50 to 70 years or older.
The elderly in low- and middle-in-
come countries experience the great-
est loss of healthy life-years due to the
non-communicable diseases affected
by PM2.5. Among those 70 years or older,
for example, PM2.5-attributable ischemic
heart disease alone accounted for 16.2% of
DALYs in China, 17.8% in India, and more
than 20% in parts of sub-Saharan Africa
and of North Africa and the Middle East
in 2016. And the air pollution-attributable
burden from non-communicable diseases is
rising, driven by high levels of air pollution
and populations that are growing and aging.
Over the past 25 years, this burden in low-
and middle-income countries has increased
for those ages 50 to 69 years — in India,
for example, increasing by 24%, from 9.1
million to 11.3 million DALYs.
These numbers have far reaching
implications. Reducing the burden of
non-communicable disease in the aging
populations of low- and middle-income
countries is among the major challenges
facing national governments and public
health officials. In 2014 the Council on
Foreign Relations noted that cardiovascu-
lar disease and other non-communicable
diseases “…killed eight million people
before their sixtieth birthdays in [low- and
middle-income countries].…The econom-
ic costs in terms of both healthcare costs
and lost productivity…[threaten] their
continued economic development and
prosperity.” Improvements in treatment
for those already suffering from cardio-
vascular disease and other non-commu-
nicable diseases will play a key role, but
a strategy of prevention of new cases of
non-communicable diseases by reducing
exposure to major risk factors such as
air pollution, high blood pressure, and
tobacco smoking will need to be aggres-
sively pursued.
and respiratory disease and lung cancer and reduced life expectan-
cy from long-term exposure to ambient air pollution. Systematic re-
views of this literature have been undertaken by organizations such
as the U.S. Environmental Protection Agency (U.S. EPA), the WHO,
and the International Agency for Research on Cancer (IARC), among
others (see “Additional Resources” at the end of this report). The
GBD project determined that the body of evidence was sufficient to
conclude that causal relationships exist between several diseases or
causes of death and exposure to ambient PM2.5, namely:
• ischemic heart disease,
• cerebrovascular disease (ischemic stroke and hemorrhagic
stroke),
• lung cancer,
• chronic-obstructive pulmonary disease (COPD), and
• lower-respiratory infections (LRIs).
Based on the current literature, for health effects attributable
to ozone, the GBD analysis includes only COPD (see “Additional
Resources”).
Our current estimates do not include causes of death and dis-
ability for which evidence for a causal relationship with exposure to
ambient PM2.5 is growing. These effects include the development of
asthma in children, low birth weight and pre-term birth, type 2 diabe-
tes, and neurological disorders. Type 2 diabetes is currently the 5th-
ranked cause of death and the 3rd-ranked cause of years lived with
disability, a component in estimating DALYs (see textbox, “Under-
standing the Burden of Disease: Deaths and DALYs”) in those aged
50 to 69 years. Multiple studies have reported that exposure to PM2.5
air pollution is a contributor to increased risk of type 2 diabetes. Re-
cent studies also suggest an association between air pollution expo-
sure and increased risk of Alzheimers disease, the 3rd-ranked global
cause of death in those 70 years and older. To date, the evidence for
these other health effects has not been judged conclusive enough to
justify their inclusion in the GBD health burden estimates, but it will
continue to be evaluated for future GBD estimates.
To estimate the potential effects of exposure to household air
pollution from the use of solid fuels for cooking or heating, scien-
tists rely on evidence both from studies of household air pollution
and from the broader air pollution and smoking literature. Studies
of exposure to household air pollution have documented impacts on
acute LRIs, COPD, cataracts, and lung cancer. Some of the strongest
evidence is for cancer; IARC has classified indoor burning of coal as
a “known human carcinogen” (IARC Group 1) and indoor burning of
biomass as a “probable human carcinogen” (IARC Group 2A). Very
few studies have evaluated household air pollution in direct relation
to ischemic heart disease or stroke, both of which are major contrib-
utors to disease burden from ambient air pollution, although a few
studies have evaluated the relation between household air pollution
11 STATE OF GLOBAL AI R / 2 0 1 8
exposure and potential indicators (e.g., blood pressure) of cardiovas-
cular disease. However, under the assumption that particles from all
combustion sources (ambient air pollution, household air pollution,
secondhand smoking, and tobacco smoking) are harmful, the GBD
project does estimate the disease burden attributable to household
air pollution exposure from ischemic heart disease and stroke using
the same integrated exposure–response relationships that it devel-
oped for ambient air pollution (see “Additional Resources”).
LEVELS AND TRENDS IN THE BURDEN OF DISEASE
FROM AMBIENT PM 2.5 FOR 2016
In 2016, long-term exposure to ambient PM2.5
contributed to 4.1 million deaths and to a loss of
106 million DALYs worldwide. China (26%) and India
(25%) together continue to bear most (51%) of the
mortality burden attributable to PM2.5.
In 2016, long-term exposure to ambient PM2.5 contributed to 4.1
million deaths and to a loss of 106 million DALYs, making PM2.5 expo-
sure responsible for 7.5% of all global deaths and 4.4% of all global
DALYs. The highest mortality burden was concentrated in Asia (Fig-
ure 9). China and India had the highest numbers of deaths attrib-
utable to PM2.5. Together, these two countries accounted for 51%
of the total global PM2.5-attributable deaths and 50% of the DALYs.
While long-term exposure to ambient PM2.5 was linked to 7.5% of all
deaths globally in 2016, it was linked to even higher percentages of
global deaths from COPD, ischemic heart disease, stroke, lung can-
cer, and LRIs (see left panel of Figure 10). Of the 4.1 million deaths
attributable to PM2.5, most (58%) were caused by ischemic heart dis-
ease and stroke (data not shown).
Overall, while the absolute numbers of deaths attributable to PM2.5
increased by over 20% from 1990 to 2016 and by 9% from 2010 to 2016
globally, there was a 12% decrease in the rate of deaths attributable
to PM2.5 from 1990 to 2016, indicating that the increase in total deaths
during this period is largely due to changes in population character-
istics. From 2010 to 2016 there was a 2.6% increase in the number
of global deaths, reflecting a greater impact of population aging and
increased global exposure than of decreases in the rates for diseases
affected by air pollution. The global trend in PM2.5-attributable deaths
broadly reflects a balance between different trends in high-income and
in low- and middle-income countries.
The global numbers and trends in PM2.5-attributable mortality,
while intriguing, can obscure important regional and country-level
patterns in individual countries that are likely to be more relevant to
local scientists and policy makers interested in understanding and
addressing the health burden attributable to poor air quality.
Figure 9. Numbers of deaths attributable to ambient PM2.5 in 2016.

Explore the data on the State of Global Air interactive site. For country abbreviations, see ISO3 website.

LEVELS AND TRENDS IN BURDEN OF DISEASE FROM Figure 10. Percentage of global deaths by cause
attributable to ambient particulate matter and to
HOUSEHOLD AIR POLLUTION FOR 2016 household air pollution.
Exposure to household air pollution from burning of
solid fuels was responsible for 2.6 million deaths and
77 million DALYs (4.7 % and 3.2% of the global totals,
respectively) in 2016, with the highest burdens in
Asia and sub-Saharan Africa.

Household air pollution was responsible for 2.6 million deaths

(4.7% of the global total) and 77 million DALYs (3.2% of the global
total) in 2016. The regional patterns of mortality burden reflect the
size of populations as well as the proportions of households using
solid fuels. China, India and its neighbors Pakistan and Bangladesh,
Indonesia, and the African nations of Nigeria, the Democratic Re-
public of Congo, Ethiopia, and Tanzania are among those countries
bearing the heaviest burden (Figure 11). The declining proportion of households using solid fuels in many
Although representing 4.7% of all deaths globally, household air countries, discussed earlier, has been reflected in decreases in
pollution exposure accounts for substantially greater percentages of global mortality attributable to household air pollution — a 30%
the worldwide mortality due to ischemic heart disease, stroke, lung decrease from 1990 (3.7 million deaths) to 2016 (2.6 million deaths).
cancer, LRIs, and COPD (see right panel of Figure 10). Of the 2.6 mil- A 16% decrease occurred in just the last 6 years. Declining death
lion deaths attributable specifically to household air pollution, 46% rates from LRIs, COPD, and stroke — in part, from other factors such
were linked to ischemic heart disease and stroke (data not shown). as improved medical care — also contribute to these trends. With

12 STATE OF GLOBAL AI R / 2 0 1 8
Figure 11. Numbers of deaths attributable to household burning of solid fuels in 2016.
Explore the data on the State of interactive site. For country abbreviations, see ISO3 website.
few exceptions, these decreases in deaths attributable to house-
hold air pollution were experienced throughout the world, with the
largest reductions in the rate of attributable deaths in North Africa,
the Middle East, and parts of Asia (e.g., South Korea, Japan, and
Malaysia).
Age-standardized DALY rates (DALYs/100,000 people) allow di-
rect comparisons between populations of different sizes and age
Figure 12. Trends in the DALY rates (DALYs per 100,000
population) attributable to household air pollution in countries
with greater than 50 million population and at least 10% of
their population using solid fuels.
13 STATE OF GLOBAL AI R / 2 0 1 8
structures. And, as the trends in Figure 12 of DALY rates attribut-
able to household air pollution indicate, in all of the countries shown
there were sharp declines, reflecting the combination of decreasing
exposure and decreases in age-standardized rates of diseases af-
fected by household air pollution, especially acute LRIs in children.
WHO BEARS MOST OF THE BURDEN FROM EXPOSURES
TO AIR POLLUTION?
Mainly because of age-related differences in mortality from chronic
diseases, exposures to ambient and household air pollution can par-
ticularly affect the very old and the very young. Of the total burden
from years of healthy life lost (DALYs), 24% was accounted for by
those over 70 years old, and 18% was associated with those less
than 5 years old. Sixty-two percent of the overall burden attributable
to ambient PM2.5 falls on those aged 50 and older. In contrast to the
findings for ambient air pollution, the burden of disease attributable
to household air pollution falls somewhat more on the young, who
already experience higher rates of lower-respiratory disease, partic-
ularly in regions were household burning of solid fuels is more prev-
alent. Nearly 28% of total DALYs for those under 5 years is attribut-
able to household air pollution exposures, compared with about 19%
Figure 13. Comparison of the percentages of deaths attributable to PM2.5,
household, and total air pollution by sociodemographic (SDI) index.
for the population older than 70 years. (For more details, see textbox,
“The Growing Burden of Disease Due to Ambient Air Pollution Expo-
sure in an Aging World” on page 10.)
COMBINED BURDEN OF EXPOSURE TO AIR POLLUTION
Since many populations may be exposed to both household and
ambient air pollution (PM2.5 and ozone), the GBD project estimates a
“total” air pollution health burden that takes them all into account.
Within the GBD analysis, the effects of exposure to household and
ambient air pollution are assumed to be largely independent, giv-
en the absence of empirical studies of their joint effects. However,
because the exposures overlap to some degree (e.g., household pol-
lution contributes to ambient pollution), the number of deaths attrib-
utable to both exposures combined is less than the sum of the deaths
due to each exposure (see Additional Resources).
Taken together, air pollution from ambient particulate
matter, ambient ozone, and household burning of solid
fuels was estimated to contribute to 6.1 million deaths
— about 11% of the total global deaths in 2016.
The estimated combined toll from all forms of air pollution (PM2.5,
ozone, and household) can be substantial. On a global basis, total
air pollution was responsible for 6.1 million deaths (11.2% of the
global total) and for 163 million DALYs (6.8% of the global total). Of
the number of global deaths in specific disease categories, 22.6%
of ischemic heart disease deaths, 21.4% of stroke deaths, 23.5% of
lung cancer deaths, 45.1% of deaths from acute LRIs, and 44.7% of
COPD deaths were attributed to air pollution in 2016.
14 STATE OF GLOBAL AI R / 2 0 1 8
Strong relationships exist between the degree
of development in and demographic standing of
a country (as measured by the SDI) and exposure
to household air pollution: countries at the lower
end of the SDI scale experience higher exposures
to household air pollution. The relation between
ambient PM2.5 exposure and sociodemographic
standing is more complex, with the highest expo-
sures found in low- and mid-SDI locations (Figure
13). Consequently, the combined impact of air
pollution is highest in those countries at “low/
middle” SDI where there is high exposure to both
risk factors — household and ambient air pollu-
tion. These countries face a “double burden” from
household air pollution on the one hand, and from
ambient PM2.5 and ozone on the other.
The combined burden from both ambient and
household air pollution falls most heavily on low-
and middle-income countries where exposures to
both are high. These countries face a double burden.
These SDI-level differences in numbers of attributable deaths
reflect differences among countries in the underlying health of the
populations, including in the fraction of total deaths accounted for by
diseases affected by air pollution, in their population’s age structure,
and in exposure levels. Consequently, comparing age-standardized
rates of death or DALYs can be more useful in comparing the burden
of disease attributable to a particular risk factor across countries or
regions, because they factor in population age structure and size.
Figure 14 provides a global map of the age-standardized death rates
from total air pollution in 2016. It shows that the combined effect of
air pollution is highest throughout much of sub-Saharan Africa and
South Asia. These patterns reflect an interaction between the levels
of exposure to the various forms of air pollution and the high rates of
diseases affected by air pollution in these regions. Further compari-
sons of age-standardized rates for individual risk factors may reveal
more information about the relative importance of those risk factors
in individual countries.
The interactive website allows for detailed comparisons of both
exposures to and burden (deaths, DALYS, and their respective rates)
attributable to air pollution between countries and among various re-
gional groupings (e.g., WHO Regions, GBD Regions, and SDI among
others) for the years 1990 to 2016.
Figure 14. Comparison of the global patterns of age-standardized death rates for both sexes attributable to
total air pollution.

Explore the data on the State of Global Air interactive site. For country abbreviations, see ISO3 website.

15 STATE OF GLOBAL AI R / 2 0 1 8
THE KEY TO AIR QUALITY PROGRESS: UNDERSTANDING THE MAJOR
SOURCES OF AIR POLLUTION
T
o reduce the disease burden attributable to air pollution
in all its forms requires a well-founded understanding of
its major contributing sources, both now and in the future.
This is the critical first step toward identifying the highest
priority actions for air quality emissions control and the
most cost-effective solutions to protect public health.
A wide range of industrial, residential, and
transportation air pollution sources contribute to
PM exposure, all needing to be controlled to reduce
health burden.
To meet that need for credible estimates of source-specific bur-
dens, HEI launched the Global Burden of Disease from Major Air
Pollution Sources (GBD MAPS) project, designed to engage local ex-
perts in China and India in (1) producing the most accurate estimates
of emissions from each source, (2) using sophisticated air pollution
models to estimate the contribution of each source to ambient PM2.5
exposure, both today and in future scenarios, and (3) applying the
GBD methods to estimate source-specific health burdens. GBD MAPS
is a multiyear collaboration among HEI, IHME, Tsinghua University,
IIT Bombay, the University of British Columbia, and other leading ac-
ademic centers.
The first major GBD report was completed in 2016 with publi-
cation of Burden of Disease Attributable to Coal-Burning and Other
Major Sources of Air Pollution in China (GBD MAPS Working Group
2016). That work has now been joined by publication in 2018 of The
Burden of Disease Attributable to Major Air Pollution Sources in In-
dia (GBD MAPS Working Group 2018). For each country, local experts
worked with the international team to assess what the sources of air
pollution health burden were for a baseline year and what they could
be in the future according to different policy pathways (“scenarios”).
CHINA
The China study found that coal-burning — by industry, by power
plants, and for residential heating — was the most important con-
tributor to ambient air pollution. Coal-burning from these three sec-
** Note that the sum of the deaths attributed to each of the coal sources
does not add up to the simulated numbers of deaths attributable to all
coal sources. This difference results both from the nonlinearities in the
relationships between emissions and ambient concentrations and from
the sensitivity simulations used to estimate the fractional contribution
from each source.
16 STATE OF GLOBAL AI R / 2 0 1 8
Figure 15. Source contributions to deaths
attributable to PM2.5 in China in 2013.
tors combined accounted for 40% of population-weighted ambient
PM2.5 concentrations and an estimated 366,200 deaths** in 2013,
the baseline year for this study (see bar graph in Figure 15).
In China, the industrial sector, comprising both coal and noncoal
sources (with 155,000 attributable deaths and 95,000 attributable
deaths, respectively), was one of the largest contributors to mortal-
ity attributable to ambient PM2.5, accounting for about 27%. How-
ever, household burning of both coal and biomass (e.g., wood and
agricultural waste) was also an important contributor to the burden
of disease attributable to ambient PM2.5. Household combustion of
these solid fuels had a combined impact (177,000 deaths) in 2015
that was larger than that from transportation (137,000 deaths) and
other sectors. Under four future policy scenarios designed to evalu-
ate the impacts of different levels of energy use and pollution control
in China, the analysis indicated that, even under the most stringent
policy scenarios, coal was projected to remain the single largest
source contributor to ambient PM2.5 and health burden in China in
2030. The findings demonstrated the air quality and health benefits
associated with continued aggressive strategies to reduce emissions
from coal combustion, along with reductions in emissions from other
major sources.
INDIA
In India, the GBD MAPS study also found significant burdens from
all forms of combustion, albeit with a different mix of the most im-
portant sources (Figure 16). In 2015, the baseline year for this study,
particulate matter air pollution from a range of major sources was
responsible for approximately 1.1 million deaths, 10.6% of the total
number of deaths in India.
Figure 16. Source contributions to deaths attributable to PM2.5 jected to be reduced in a major way — by nearly 35%
in India in 2015. from 2015 to 2050, reaching about 48 µg/m3. Even with
the projected exposure decreases, the burden of disease
is expected to grow in the future relative to the baseline
year as the population ages and grows and leaves more
people susceptible to air pollution.
The Indian government has begun taking actions to
improve air quality. However, the study reported, the
most aggressive action — with all major sectors achiev-
ing reductions in air pollution — could avoid up to 1.2
million deaths in 2050 compared with just instituting
currently planned policies. Achieving such reductions
will require particular attention to reducing emissions
from household biomass combustion, coal burning, and
dusts related to human activities. Depending on the sce-
nario, coal combustion has the potential to emerge as
the leading contributor to the disease burden from air
pollution in the future.

• Residential biomass burning was the largest individual

SUMMARY
contributor to disease burden. Residential biomass burn-
ing was responsible for 267,700 deaths or nearly 25% of the These two studies have demonstrated that the reduction of air
deaths attributable to PM2.5, making it the most important sin- pollution exposure and health burden in China and India — and
gle anthropogenic source related to mortality in 2015. These likely many other countries — will require continued and updated
burden estimates do not include the considerable additional source-specific analyses as actions are taken to measure progress
burden from indoor exposure to biomass burning in the home. and identify continuing challenges. The GBD MAPS initiative is cur-
• Coal combustion and open burning also contributed rently putting in place the next generation of such analyses for China
substantially to disease burden. Coal combustion, roughly and India, with other countries to follow.
evenly split between industrial sources and thermal power
plants, was responsible for 169,300 deaths (15.5%) in 2015.
The open burning of agricultural residue was responsible for
66,200 (6.1%) PM2.5-attributable deaths.
• Transport, distributed diesel, and brick production
were also important contributors to PM2.5-attributable
disease burden. In 2015, transportation contributed 23,100
deaths, distributed diesel contributed 20,400 deaths and brick
production contributed 24,100 deaths.

The study also reported that if no further action is taken, popula-

tion exposures to PM2.5 are likely to increase by over 40% by 2050.
Three different energy efficiency and air pollution control scenarios
were evaluated. Only under the most aspirational scenario, in which
aggressive emission reductions were assumed, were exposures pro-

17 STATE OF GLOBAL AI R / 2 0 1 8
CONCLUSIONS
T
he GBD project has played a key role in identifying the
factors that contribute the most to disease and prema-
ture mortality — the first step toward determining what
can be done to improve public health. Among the 84 risk
factors included in its comprehensive analysis, the GBD
project reported that ambient air pollution from PM2.5
ranked 6th globally in its contribution to mortality in 2016, account-
ing for 4.1 million deaths. Household air pollution was ranked 8th
globally, responsible for 2.6 million deaths. Air pollution from ambi-
ent PM2.5, ozone, and household burning of solid fuels combined was
the 4th-highest global risk factor, accounting for 6.1 million deaths
— 11% of the global total. The GBD analysis has also laid out the
critical interplay between the trends in population structure, under-
18 STATE OF GLOBAL AI R / 2 0 1 8
lying disease, and economic factors and the trends in air pollution
levels. Knowledge of these trends is essential to understanding pat-
terns in the burden of disease experienced by different countries and
regions and to helping to inform decision makers where policy action
to reduce population exposure at the national or regional levels has
the most potential to provide large benefits in improved health. Ul-
timately, reduction in air pollution and its burden on health requires
identifying and taking action to control the major sources that con-
tribute to them. Actions to reduce air pollution should address not
only the larger-scale burning of coal by power plants and industries,
but also the use of coal or different forms of biomass for heating and
cooking in millions of small households around the world.
 Numbers, Numbers Everywhere
As recognition of the world’s air pollution problems has grown,
estimates of the numbers of deaths and years of healthy life lost
attributable to outdoor air pollution have proliferated. Most of
these health burden estimates are from the World Health Organi-
zation (WHO) or from the Institute for Health Metrics and Evalua-
tion (IHME) Global Burden of Disease (GBD) project. (This State of
Global Air report is based on IHME estimates.)
In its most recent update, released in 2016, the WHO estimated
there were 3 million deaths from PM2.5 exposure for the year 2012,
while the most recent GBD estimate was 4.1 million for the year
2016. Other estimates exist for outdoor air pollution–related deaths
in individual countries or regions, alone and at times in combination
with estimates for household air pollution. Indeed, the GBD 2016
estimates 6.1 million deaths attributable to all forms of air pollution:
4.1 million from PM2.5, 2.6 million from household, and the remain-
der from ozone. Over time, these or related numbers have been
echoed in reports by several leading economic and energy institu-
tions that have sought to put a monetary value on the health burden
attributable to outdoor air pollution and its sources — the World
Bank, the Organization for Economic Cooperation and Development,
and the International Energy Agency, to name a few (see “Additional
Resources” section for details).
WHAT CAN WE MAKE OF ALL THESE NUMBERS? DO
THEIR DIFFERENCES MATTER?
The most important takeaway message is that they are all large
numbers — the burden of disease from air pollution is substan-
tial. And given the complexity of the process for developing
them, these estimates are surprisingly consistent. Some varia-
tion from this kind of scientific analysis is to be expected. These
are estimates made by different analysts at different points in
time. They vary primarily because of different data or because
of different methods used to assess exposure to pollutants, to
characterize exposure–disease relationships, and to quantify the
baseline rates of disease and mortality in populations — all of
which go into estimates of the numbers of people affected by air
pollution.
Increasingly, the methods for estimating the burden of disease
are converging. Most of the data and methods have their origins
19 STATE OF GLOBAL AI R / 2 0 1 8
with IHME’s GBD project and, particularly, its major update in 2010,
which substantially expanded the analyses. WHO and IHME now
use essentially the same methodologies for estimating air pollution
levels around the world. Their estimates are therefore most sensi-
tive to changes in exposure–response relationships that occur as
new evidence is incorporated and, to a lesser degree, to changes in
the baseline disease and mortality rates. Differences in the choice of
exposure–response relationships largely account for the differences
between the most recent mortality estimates from WHO and GBD:
WHO relied on exposure–response relationships from GBD 2013
and WHO population mortality rate estimates for 2012, whereas the
GBD 2016 report used an updated exposure–response function that
predicted higher rates of mortality from PM2.5 exposure and 2016
estimates of population mortality rates.
ANNUAL GBD UPDATES
The GBD project now updates its estimates annually and, with
each update, provides an analysis of the trends over time (e.g.,
for the 26 years from 1990 to 2016). Although these updates
include improvements in data and methods that themselves
contribute to differences from previous GBD estimates (e.g., GBD
2010, GBD 2013, and GBD 2015), each GBD update recalculates
the entire temporal sequence so that its trends (based, e.g., on
the years 1990–2016) are internally consistent. These are the
data that will be featured in future State of Global Air reports.
PUTTING IT ALL TOGETHER — A GLOBAL ESTIMATE
OF POLLUTION’S IMPACT
The Lancet Commission on Pollution and Health has recently
used the GBD methods and data to estimate the disease burden
from all forms of pollution — ambient, household air, water, soil,
heavy metal, and chemical pollution. The Commission estimated
that 9 million premature deaths, about 16% of the total global
mortality, could be attributed to pollution-related disease in
2015 (see Additional Resources). The World Health Organization,
using additional definitions of pollution-related risk factors, had
put the number at 12.6 million in 2012. Again, the size of these
numbers is more important than their differences.

ADDITIONAL RESOURCES
GBD 2016 METHODS
Details on the methods used to estimate PM2.5, ozone, and house-
hold air pollution exposures and to estimate the premature deaths
and disability-adjusted life-years (DALYs) and their respective rates
for the GBD 2016 analyses can be found in these studies and their
related references:
Cohen AJ, Brauer M, Burnett R, Anderson HR, Frostad J, Estep K, et
al. 2017. Estimates and 25-year trends of the global burden of dis-
ease attributable to ambient air pollution: An analysis of data from
the Global Burden of Diseases Study 2015. Lancet 389:1907–1918;
http://dx.doi.org/10.1016/S0140-6736(17)30505-6.
GBD 2016 Risk Factors Collaborators. 2017. Global, regional, and
national comparative risk assessment of 84 behavioural, envi-
ronmental and occupational, and metabolic risks or clusters of
risks, 1990–2016: A systematic analysis for the Global Burden of
Disease Study 2016. Lancet 390:1345–1422; doi: 10.1016/S0140-
6736(17)32366-8.
Shaddick G, Thomas M, Jobling A, Brauer M, van Donkelaar A,
Burnett R, et al. 2016. Data integration model for air quality: A hi-
erarchal approach to the global estimation of exposures to ambient
air pollution. Available: https://arxiv.org/abs/1609.00141 [accessed
9 February 2018].
At the IHME GBD Compare website, data on mortality and disease
burden for air pollution, as well as other risk factors, may be further
explored and downloaded.
HEALTH EFFECTS OF PM 2.5, OZONE, AND HOUSEHOLD
AIR POLLUTION
For comprehensive reviews of the scientific evidence on the health
effects associated with exposures to PM2.5 and ozone from the U.S.
Environmental Protection Agency and the WHO, see the following:
Daniels, ME, Jr, Donilon TE, Bollyky TJ. 2014. The Emerging Global
Health Crisis: Noncommunicable Diseases in Low- and Middle-In-
come Countries. Independent Task Force Report No. 72. New York,
NY:Council on Foreign Relations. Available: https://cfrd8-files.cfr.
org/sites/default/files/report_pdf/TFR72_NCDs.pdf [accessed 23
February 2018].
20 STATE OF GLOBAL AI R / 2 0 1 8
<<Return
Health Effects Institute. In press. Household Air Pollution and
Non-Communicable Diseases. Communication 18. Boston, MA:
Health Effects Institute.
IARC (International Agency for Research on Cancer). 2013. Air
Pollution and Cancer. IARC Scientific Publication No. 161. Lyon,
France:World Health Organization. Available: www.iarc.fr/en/pub-
lications/books/sp161/AirPollutionandCancer161.pdf [accessed 8
February 2018].
U.S. EPA (United States Environmental Protection Agency). 2013.
Final Report: Integrated Science Assessment (ISA) for Ozone and
Related Photochemical Oxidants (Final Report, Feb 2013). EPA/600/
R-10/076F, 2013. Washington, DC:U.S. Environmental Protection
Agency. Available: https://cfpub.epa.gov/ncea/isa/recordisplay.
cfm?deid=247492 [accessed 12 February 2018].
U. S. EPA (United States Environmental Protection Agency). 2009.
Final Report: Integrated Science Assessment (ISA) for Particulate
Matter (Final Report, Dec 2009). EPA/600/R-08/139F, 2009. Wash-
ington, DC:U.S. Environmental Protection Agency. Available: http://
cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=216546 [accessed 8
February 2018].
WHO (World Health Organization). 2015. Air Quality Guidelines:
Global Update 2005. WHO Reg Off Eur. Available: www.who.int/
phe/health_topics/outdoorair/outdoorair_aqg/en/ [accessed 12
February 2018].
WHO (World Health Organization). 2014. Burden of Disease from
Household Air Pollution for 2012. Geneva, Switzerland:World Health
Organization. Available: www.who.int/phe/health_topics/outdoo-
rair/databases/HAP_BoD_results_March2014.pdf?ua=1 [accessed
21 February 2018].
WHO (World Health Organization). 2016. Burning Opportunity:
Clean Household Energy for Health, Sustainable Development, and
Wellbeing of Women and Children. Geneva, Switzerland:World
Health Organization. Available: http://apps.who.int/iris/bitstre
am/10665/204717/1/9789241565233_eng.pdf [accessed 21 Febru-
ary 2018].
WHO (World Health Organization). 2014. WHO Guidelines for Indoor
Air Quality: Household Fuel Combustion. Executive Summary. Gene-
va, Switzerland:World Health Organization. Available: http://apps.
who.int/iris/bitstream/10665/144309/1/WHO_FWC_IHE_14.01_
eng.pdf?ua=1 [accessed 21 February 2018].
GBD MAPS
GBD MAPS Working Group. 2016. Burden of Disease Attributable
to Coal-Burning and Other Major Sources of Air Pollution in China.
Special Report 20. Boston, MA:Health Effects Institute.
GBD MAPS Working Group. 2016. Executive Summary. Burden of
Disease Attributable to Coal-Burning and Other Major Sources of
Air Pollution in China. Special Report 20. Boston, MA:Health Effects
Institute.
GBD MAPS Working Group. 2018. Burden of Disease Attributable
to Major Air Pollution Sources in India. Special Report 21. Boston,
MA:Health Effects Institute.
GBD MAPS Working Group. 2018. Summary for Policy Makers. Bur-
den of Disease Attributable to Major Air Pollution Sources in India.
Special Report 21. Boston, MA:Health Effects Institute.
HOUSEHOLD AIR POLLUTION
Clark ML, Peel JL, Balakrishnan K, Breysse PN, Chillrud SN, Naeher
LP, et al. 2013. Health and household air pollution from solid fuel
use: The need for improved exposure assessment. Environ Health
Perspect 121:1120–1128; doi: 10.1289/ehp.1206429.
GBD 2016 Risk Factors Collaborators. 2017. Global, regional, and
national comparative risk assessment of 84 behavioural, envi-
ronmental and occupational, and metabolic risks or clusters of
risks, 1990–2016: A systematic analysis for the Global Burden of
Disease Study 2016. Lancet 390:1345–1422; doi: 10.1016/S0140-
6736(17)32366-8.
GBD 2015 Risk Factors Collaborators. 2016. Global Burden of
Disease Study 2015 (GBD 2015) Socio-Demographic Index (SDI)
1980–2015. Available: http://ghdx.healthdata.org/record/global-
burden-disease-study-2015-gbd-2015-socio-demographic-index-sdi-
1980%E2%80%932015 [accessed 14 February 2018].
World Bank and Institute for Health Metrics and Evaluation.
2016. The Cost of Air Pollution: Strengthening the Economic Case for
Action. Washington, D.C.:World Bank Group. Available: http://docu-
ments.worldbank.org/curated/en/781521473177013155/The-cost-of-
air-pollution-strengthening-the-economic-case-for-action [accessed
16 December 2016].
21 STATE OF GLOBAL AI R / 2 0 1 8
NUMBERS, NUMBERS EVERYWHERE
Listed below are the reports and papers referred to in the textbox
“Numbers, Numbers Everywhere”:
IARC (International Energy Agency). 2016. Energy and Air Popu-
lation: World Energy Outlook Special Report. Paris, France:Inter-
national Energy Agency. Available: www.iea.org/publications/
freepublications/publication/WorldEnergyOutlookSpecialReport-
2016EnergyandAirPollution.pdf [accessed 12 February 2018].
Landrigan PJ, Fuller R, Acosta NJR, Adeyi O, Arnold R, Basu N, et
al. 2017. The Lancet Commission on Pollution and Health. Lancet
391:462–512; https://doi.org/10.1016/S0140-6736(17)32345-0.
Organization of Economic Cooperation and Development. 2014. The
Cost of Air Pollution: Health Impacts of Road Transport. Paris:OECD
Publishing. Available: http://dx.doi.org/10.1787/9789264210448-en
[accessed 12 February 2018].
WHO (World Health Organization). 2016. Ambient Air Pollution:
A Global Assessment of Exposure and Burden of Disease. Ge-
neva:World Health Organization. Available: http://apps.who.int/
iris/bitstream/10665/250141/1/9789241511353-eng. pdf?ua=1
[accessed 12 February 2018].
WHO (World Health Organization). 2016. WHO Global Urban Ambi-
ent Air Pollution Database (Update 2016). Available: http://www.who.
int/phe/health_topics/outdoorair/databases/cities/en/ [accessed 12
February 2018].
World Bank and Institute for Health Metrics and Evaluation. 2016.
The Cost of Air Pollution: Strengthening the Economic Case for
Action. Washington, DC:World Bank Group. Available: http://
documents.worldbank.org/curated/en/781521473177013155/The-
cost-of-air-pollution-strengthening-the-economic-case-for-action
[accessed 12 February 2018].
CONTRIBUTORS AND FUNDING
CONTRIBUTORS
Health Effects Institute: HEI is an independent global health and
air pollution research institute, a leader of the air pollution analysis
within the Global Burden of Disease (GBD) project, and the producer,
most recently, of the GBD from Major Air Pollution Sources (GBD
MAPS) reports on China and India. HEI is the primary developer of the
State of Global Air report, host and manager for the related website,
coordinator of input from all other members of the team, and facilita-
tor of contact with media partners.
The Institute for Health Metrics and Evaluation: IHME is
an independent global health research center at the University of
Washington, which provides rigorous and comparable measurement
— through the GBD project — of the world’s most important health
problems and the risk factors that contribute to them, and which
evaluates the strategies used to address them. A key collaborator,
IHME provides the underlying air pollution and health data and other
critical support for this project.
University of British Columbia: Professor Michael Brauer of
the School of Population and Public Health at UBC is a critical exter-
nal expert advising on this project. Dr. Brauer is a long-time principal
collaborator on the air pollution assessment for the GBD project and
led the effort to define the project’s global air pollution exposure as-
sessment methodology.
22 STATE OF GLOBAL AI R / 2 0 1 8
ADDITIONAL ACKNOWLEDGMENTS
The State of Global Air project would not have been possible with-
out major contributions by HEI staff: Katy Walker, Principal Scientist;
Hilary Selby Polk, Managing Editor; Annemoon van Erp, Managing
Scientist; and Kathryn Liziewski, Research Assistant. We also ac-
knowledge the contribution of Aaron Cohen, a long-time Principal
Scientist at HEI and now Affiliate Professor of Global Health at IHME,
who has played a leading role in the development of the GBD air
pollution estimates.
We are indebted to Zev Ross, Hollie Olmstead, and Chris Marx
at ZevRoss Spatial Analysis for their development of the interactive
feature of the website.
The State of Global Air website was designed by Glenn Ruga at
Glenn Ruga Visual/Communications and built by Ezra Klughaupt and
Diane Szczesuil at Charles River Web.
We thank Ajay Pillarsetti for his photographs of household air
pollution.
FUNDING
The State of Global Air project has been funded by the William and
Flora Hewlett Foundation and the Oak Foundation.
Glossary
For a glossary of terms, see the State of Global Air website.
How to Cite This Publication
Health Effects Institute. 2018. State of Global Air 2018. Special
Report. Boston, MA:Health Effects Institute.
Health Effects Institute
75 Federal Street, Suite 1400
Boston, MA 02110, USA